Faculty Bibliography

We have constructed the first comprehensive microarray representing a human chromosome for analysis of DNA copy number variation. This chromosome 22 array covers 34.7 Mb, representing 1.1% of the genome, with an average resolution of 75 kb. To demonstrate the utility of the array, we have applied it to profile acral melanoma, dermatofibrosarcoma, DiGeorge syndrome and neurofibromatosis 2. We accurately diagnosed homozygous/heterozygous deletions, amplifications/gains, IGLV/IGLC locus instability, and breakpoints of an imbalanced translocation. We further identified the 14-3-3 eta isoform as a candidate tumor suppressor in glioblastoma. Two significant methodological advances in array construction were also developed and validated. These include a strictly sequence defined, repeat-free, and non-redundant strategy for array preparation. This approach allows an increase in array resolution and analysis of any locus; disregarding common repeats, genomic clone availability and sequence redundancy. In addition, we report that the application of phi29 DNA polymerase is advantageous in microarray preparation. A broad spectrum of issues in medical research and diagnostics can be approached using the array. This well annotated and gene-rich autosome contains numerous uncharacterized disease genes. It is therefore crucial to associate these genes to specific 22q-related conditions and this array will be instrumental towards this goal. Furthermore, comprehensive epigenetic profiling of 22q-located genes and high-resolution analysis of replication timing across the entire chromosome can be studied using our array.

Aberrant methylation of CpG islands and genomic deletion are two predominant mechanisms of gene inactivation in tumorigenesis, but the extent to which they interact is largely unknown. The lack of an integrated approach to study these mechanisms has limited the understanding of tumor genomes and cancer genes. Restriction landmark genomic scanning (RLGS; ref. 1) is useful for global analysis of aberrant methylation of CpG islands, but has not been amenable to alignment with deletion maps because the identity of most RLGS fragments is unknown. Here, we determined the nucleotide sequence and exact chromosomal position of RLGS fragments throughout the genome using the whole chromosome of origin of the fragments and in silico restriction digestion of the human genome sequence. To study the interaction of these gene-inactivation mechanisms in primary brain tumors, we integrated RLGS-based methylation analysis with high-resolution deletion maps from microarray-based comparative genomic hybridization (array CGH; ref. 3). Certain subsets of gene-associated CpG islands were preferentially affected by convergent methylation and deletion, including genes that exhibit tumor-suppressor activity, such as CISH1 (encoding SOCS1; ref. 4), as well as genes such as COE3 that have been missed by traditional non-integrated approaches. Our results show that most aberrant methylation events are focal and independent of deletions, and the rare convergence of these mechanisms can pinpoint biallelic gene inactivation without the use of positional cloning.

Genomic abnormalities at 348 loci encoding genes that may contribute to lung cancer transformation and progression were assessed using array comparative genomic hybridization in 21 squamous carcinomas (SqCas) and 16 adenocarcinomas (AdCas). Hierarchical clustering showed a clear pattern of gains and losses for the SqCas, whereas the pattern for AdCas was less distinct. Cross-validated classification using a K-nearest-neighbor assigned, on average, 32 of 37 samples to their proper histological subtype. The most noticeable differences between SqCas and AdCas were gain of chromosome 3q22-q26 and loss of chromosome 3p. These occurred almost exclusively in SqCas. The region of recurrent increase is approximately 30 Mb in extent, ranging from EVI1 to TFRC. PIK3CA, the alpha catalytic subunit of phosphatidylinositol 3-kinase (PI3K), is in this region. The PIK3CA copy number increase was validated using fluorescence in situ hybridization to lung cancer tissue microarrays. Activity of the downstream PI3K effector protein kinase B (PKB) was higher in SqCas than in AdCas and was correlated with PIK3CA copy number (r = 0.75), suggesting that these copy number increases contribute to activation of PI3K signaling in SqCas of the lung.

Cardio-facio-cutaneous (CFC) syndrome is characterized by a distinct facial appearance, cardiac defects, ectodermal anomalies and developmental delay. Recently, we reported a 19-month-old girl with phenotypic manifestations consistent with the CFC syndrome who had an interstitial deletion of the long arm of chromosome 12, del(12)(q21.2q22), implicating a possible locus for CFC syndrome. Here, we report an additional patient with a cytogenetically identical interstitial deletion: 47,XYY,del(12)(q21.2q22). To further characterize this deletion we used microarray-based comparative genomic hybridization (array CGH). Array CGH confirmed both the deletion and the second Y chromosome. The deletion on chromosome 12q spanned at least 14 Mb as indicated by the positions on the genome sequence of the 4 BAC clones included in the deletion. While the proband did not have the classic features of CFC, he had some dysmorphic craniofacial characteristics, ectodermal anomalies and moderate developmental delay which were suggestive of CFC syndrome; however, this patient did not have classical CFC. The phenotypic differences between the two del(12)(q21.2q22) patients may be due to variability in the expression of the syndrome, or this deletion may present as a syndrome with overlapping features. Alternatively, the phenotypic differences may result from discordance at the molecular level, which may yield a critical minimal region of deletion for CFC. The region 12q21.2 --> q22 remains a possible candidate region for CFC syndrome. Additional characterization of these and other CFC patients may confirm and further refine this candidate region.

Various members of the Ets multigene family exhibit diverse roles in development, cell differentiation, tissue-specific gene expression and human malignancy. In the search for Ets factors involved in mammary gland development and malignancy, ESX was found to be upregulated in a subset of breast tumours and cell lines. We report the transcriptional regulation of ESX in epithelial breast cancer cells. Transient reporter assays using the ESX promoter show that ESX transcription is regulated by ErbB receptor signalling. In cell lines and in 45 primary ductal breast cancers we show that ESX transcript expression significantly correlates with ErbB2 transcript levels. Moreover, expression of ErbB2 in cells upregulates ESX promoter activity while inhibition of ErbB2 or its downstream signaling pathways decrease both ESX promoter activity and endogenous ESX protein levels. These results indicate that the ESX promoter represents a transcriptional target of ErbB2, and ESX expression may represent a downstream mediator of ErbB2 signaling and ErbB2-induced gene expression.

DNA microarrays are now widely used to measure expression levels and DNA copy number in biological samples. Ratios of relative abundance of nucleic acids are derived from images of regular arrays of spots containing target genetic material to which fluorescently labeled samples are hybridized. Whereas there are a number of methods in use for the quantification of images, many of the software systems in wide use either encourage or require extensive human interaction at the level of individual spots on arrays. We present a fully automatic system for microarray image quantification. The system automatically locates both subarray grids and individual spots, requiring no user identification of any image coordinates. Ratios are computed based on explicit segmentation of each spot. On a typical image of 6000 spots, the entire process takes less than 20 sec. We present a quantitative assessment of performance on multiple replicates of genome-wide array-based comparative genomic hybridization experiments. By explicitly identifying the pixels in each spot, the system yields more accurate estimates of ratios than systems assuming spot circularity. The software, called, runs on Windows platforms and is available free of charge for academic use.

Microarray measurements offer the potential to compare the abundances of numerous nucleic acid sequences in parallel. Using linker-adapter PCR products from mapped BAC clones we have made arrays that permit scanning the human genome for single copy gains and losses of DNA sequence, which requires reliable detection of 50% changes. The DNA is printed at high concentration on amino-silane or chromium coated surfaces using a custom-built capillary pin printing system. Spots are printed on 130 μm centers or closer to minimize the size of the arrays. Hybridization occurs in a dextran sulfate/formamide buffer at 37°C, using slow rocking to mix the reaction. The entire array is imaged in a single CCD frame using a custom built system that employs mercury arc illumination. Up to four fluorochromes can be imaged from a single array with adequate spectral separation. Typically we use DAPI to stain the DNA in the array spots to facilitate automatic image segmentation during analysis, and fluorescein, Cy3, and Cy5 or their spectral equivalents, for labeling specimen nucleic acids. Array spots are segmented and quantitative fluorescence intensities and intensity ratios are automatically calculated in < 1 minute per ∼8000 element array using the custom software UCSF SPOT.

Carcinomas that develop in the pancreatic islets of transgenic mice expressing the SV40 T-antigens (Tag) under transcriptional control of the rat insulin II promoter (RIP) progress through well-characterized stages that are similar to aspects of human tumor progression, including hyperplastic growth, increased angiogenesis and reduced apoptosis. The latter two stages have been associated with recurrent loss of heterozygosity (LOH) and reduced genome copy number on chromosomes 9 (LOH9) and 16 (LOH16), aberrations which we believe contribute to these phenotypes. Earlier analyses localized LOH9 to approximately 3 Mb and LOH16 to approximately 30 Mb (both syntenic with human 3q21-q25) but were limited by low throughput and a lack of informative polymorphic markers. Here we show that comparative genomic hybridization to DNA microarrays (array CGH) overcomes these limitations by allowing efficient, genome-wide analyses of relative genome copy number. The CGH arrays used in these experiments carried BACs distributed at 2-20-MB intervals across the mouse genome and at higher density in regions of interest. Using array CGH, we further narrowed the loci for LOH9 and LOH16 and defined new or previously unappreciated recurrent regions of copy-number decrease on chromosomes 6, 8 and 14 (syntenic with human chromosomes 12p11-p13, 16q24.3 and 13q11-q32, respectively) and regions of copy-number increase on chromosomes 2 and 4 (syntenic to human chromosomes 20q13.2 and 1p32-p36, respectively). Our analyses of human genome sequences syntenic to these regions suggest that CYP24, PFDN4, STMN1, CDKN1B, PPP2R3 and FSTL1 are candidate oncogenes or tumor-suppressor genes. We also show that irradiation and genetic background influence the spectrum of aberrations present in these tumors.

We have assembled arrays of approximately 2,400 BAC clones for measurement of DNA copy number across the human genome. The arrays provide precise measurement (s.d. of log2 ratios=0.05-0.10) in cell lines and clinical material, so that we can reliably detect and quantify high-level amplifications and single-copy alterations in diploid, polyploid and heterogeneous backgrounds.